My only concern reading this is that LLVM already has a similar mechanism for channel-based logging that is used throughout the backends. Is there a reason to invent our own? Maybe we can take what is in LLVM, enhance it and use shared functionality?
There are a few differences between your solution and the LLVM one, but I'm not sure they are sufficient to justify creating a new solution. The LLVM debug logging solution is documented in the LLVM Programmers Manual here: http://llvm.org/docs/ProgrammersManual.html#the-debug-macro-and-debug-option -Chris > On Dec 6, 2016, at 8:57 AM, Pavel Labath <lab...@google.com> wrote: > > On 6 December 2016 at 16:51, Jim Ingham <jing...@apple.com> wrote: >> Sorry, I'm being dense. What is Dxxxxx? >> >> Jim >> >>> On Dec 6, 2016, at 8:23 AM, Pavel Labath <lab...@google.com> wrote: >>> >>> I have created straw-man implementation of such an interface in >>> Dxxxxx, together with some examples of how it's used, and I'd like to >> > > https://reviews.llvm.org/D27459 :D > > Sorry, I forgot to replace this by actual revision when I submitted the code. > > pl _______________________________________________ lldb-dev mailing list lldb-dev@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-dev
